1. Field of the Invention
This invention relates in general to the field of semiconductor devices. More particularly, the invention relates to quantum-well semiconductor lasers and optical amplifiers.
2. Description of the Prior Art
Semiconductor lasers emit optical radiation in a highly directional beam as do solid-state lasers and gas lasers. However, the semiconductor laser differs from other lasers in that it is small and is easily modulated at high frequencies simply by modulating the biasing current. A semiconductor optical amplifier is a device that closely resembles a semiconductor laser. The semiconductor optical amplifier, which does not have a resonant cavity, uses its biasing current to amplify appropriately directed light incident on its active region.
Because of their small size and unique performance characteristics, the semiconductor laser and optical amplifier have found widespread use in many fields. One of the most important uses of the semiconductor laser is as a light source for optical-fiber communication. Also, semiconductor lasers and optical amplifiers have significant applications in video recording, optical signal processing, high-speed printing, and optical scanning, reading and display devices.
One common type of semiconductor laser/amplifier comprises a quantum-well heterostructure of intrinsic semiconductor material sandwiched between a highly doped P-region and a highly doped N-region. Light is emitted from the region that contains the quantum-well heterostructure, and, as such, that region is referred to as the active region. The P- and N-regions, which are called inactive regions, provide suitable carriers to the active region and act as a waveguide for the light. As is well known by those skilled in these arts, these regions can be fabricated by conventional epitaxial crystal growth techniques using, for example, III-V compound alloy systems to form the semiconductor layers. In general, these structures are referred to as PIN diode lasers/amplifiers.
To achieve heterostructures with negligible interface traps, the lattices between the different semiconductor layers must be matched closely. Large lattice mismatches usually result in the formation of lattice defects such as dislocations that degrade device performance. It has been recognized, however, that moderate mismatches that produce coherently strained layers free of dislocations can be beneficial for semiconductor laser/amplifier performance as opposed to devices without strain. Specifically, the modified band structure of strained quantum-wells often lead to significant benefits for diode laser performance, including reduced threshold current density, improved efficiency and temperature insensitivity, and enhanced dynamic response and high-speed performance. As such strained lasers and optical amplifiers are commonly available at a wide range of wavelengths. Doyeol Ahn and Shun Lien Chuang describe a theoretical model for such devices in their publication entitled "The Theory of Strained-Layer Quantum-Well Lasers with Bandgap Renormalization," IEEE Journal of Quantum Electronics, Vol. 30, No. 2, February 1994, pp 350-365.
Obtaining maximum strain through lattice mismatching without inducing dislocations can be a critical problem for semiconductor engineers. The formation of dislocations is generally a function of layer thickness and the degree of mismatch. Specifically, for a given mismatch there is a critical layer thickness above which dislocations will form. The critical layer thickness constraint normally limits the maximum strain in a practical quantum-well layer to about one percent. Beyond that one-percent strain limit, substantial dislocations form. Attempts to increase strain beyond the one percent value and avoid dislocations by reducing layer thickness typically result in producing a quantum well that is to thin to have practical applications.
Consequently, those concerned with the development of strained-layered quantum-well semiconductor lasers and optical amplifiers have long sought techniques for obtaining enhanced performance by exploiting the beneficial effects of increased strain. When engineering such devices, however, designers have encountered considerable difficulty in avoiding the formation of lattice dislocations when attempting to increase strain by increasing the degree of lattice mismatch. As mentioned above, the critical layer thickness constraint normally limits the maximum strain in a practical quantum-well layer to about one percent. The present invention mitigates the problems associated with this limitation.